There’s something rewarding about watching a simple idea turn into a working cloud app. During the AI for Bharat workshop, I built a serverless image editing app using AWS services. Here’s how everything came together.
🌈 What I Built
A web app where users can:
- Sign in securely
- Upload an image
- Request AI-powered edits
- View the final output
Powered by: Amazon Bedrock (Titan Image Generator G1)
Supported by: Cognito, DynamoDB, Lambda, API Gateway, Amplify
🧩 Architecture Overview
User Authentication (Cognito)
Handles sign-in and protects routes.
Frontend (Amplify)
Simple hosting and deployment.
API Layer (API Gateway)
Connects frontend to the backend securely.
Backend Logic (Lambda)
Processes prompts, invokes Bedrock, saves data.
Database (DynamoDB)
Stores prompts, timestamps, image details.
Final Output
Edited image returns to the frontend.
🧭 Key Learnings
🔐 Cognito
Smooth authentication setup.
🗃️ DynamoDB
Great for schema-less storage.
⚙️ Lambda
Serverless Python functions that just work.
🔌 API Gateway
Handles routing and authorization.
🎨 Amazon Bedrock
Titan G1 felt like a creative engine.
Prompts like “make it cinematic” or “turn it into anime” instantly produced results.
🌐 Amplify
Frontend deployment was fast and painless.
🧠 Skills Gained
- Serverless architecture
- IAM permissions
- NoSQL data modeling
- REST API design
- AWS Lambda development
- Bedrock AI integration
- Amplify hosting
- Full-stack cloud development
🕒 Time Taken
Approximately 60 minutes following the workshop modules.
💸 Cost
Around $2–5 depending on usage.
🖼️ Final Output
🌟 Final Thoughts
This project felt like assembling a miniature AI-powered ecosystem where every AWS service cooperated effortlessly. No server overhead, no scaling issues — just clean serverless design.
If you enjoy building with cloud technologies, this workshop is a perfect playground.
Maybe this prototype becomes something bigger one day.
Happy building! ✨
Top comments (0)